The development of additional agents for the treatment of tuberculosis is of great importance due to the persistence of this disease as a major cause of mortality in the less developed countries and the emergence of multi-drug resistant tuberculosis in HIV infection individuals (particularly prisoners and intravenous drug users) in this country. Ultra short course chemotherapy (less than 4 month regimens) and new regimens with activity comparable to isoniazid and rifampin are urgently needed. Several new agents studied in vitro and in murine models of tuberculosis have great promise. Among these are the rifamycins (rifapentine, KRM 1648 & KRM 1657), quinolones (sparfloxacin and levofloxacin), oxazolidinones and perhaps the macrolides. These agents and others will be evaluated in a dose response study using a murine tuberculosis model. The pharmacokinetics of these agents will be studied in mice to ensure that in vivo testing utilizes doses that are clinically relevant. The most promising agents will be evaluated in combination (two or three agents) against the standard isoniazid/rifampin regimen. It is feasible to utilize the murine model to develop candidate regimens for pilot studies in humans. The in vivo response to agents determined to be "resistant" in vitro will be evaluated with particular attention to isoniazid, ethionamide and rifampin. This has not been adequately studied in the past in the murine tuberculosis model. Due to the general lack of availability of MIC determinations for antituberculosis agents it is important to determine whether in vitro resistance correlates with in vivo activity in the mouse tuberculosis model.